Grid scale energy storage in dry sand

[Dave]

I've been thinking and watching videos about energy storage at grid
scale. Needed because wind power is variable, and not everyone has the
ability to store energy in their home.

Pump storage is OK for a day or two, but is limited by geography.
Liquid air is interesting, and batteries are too expensive and resource
constrained.

Thermal storage seems a good idea. Probably best to build this locally
where the heat is needed if there is the infrastructure to distribute
heat. For a year or two I stayed in a flat with district heating.
However the metering was awful - use a lot or a little, by bill was
always GBP 60/month.

So the concept is storing the heat in dry sand, and move the heating
elements or heat extraction pipes. In this way there is a decoupling
between the expensive (heat exchanger) from the scalable parts: the
building, insulation and sand.

How much? Say have a 18 meter high building, which has total trays of
sand 500m x 200 m and 10m deep. Making this 1 million cubic meters and
1.4 million tons for sand. Sand has 850 J Kg-1 K-1 thermal capacity.
Say, heat to 800 Celsius, and allow to cool to 400 Celsius so a 400
Celsius usable drop. This is 400 * 850 * 14000000 * 1000 Joules, which
is 476TJ or dividing by 3600 to get 132 GWh. Want storage for 1 week
minimum, so divide by 168 to get 0.78 GWh.

People who work in power stations know the best way to transfer the
heat- specialist fluid or superheated water. Because this is a post on
Usenet, I'm not doing an integral, so I'm assuming a midpoint for the
temperature for the efficiency of 580 Celsius. With ambient of 291K, get
maximum efficiency of 1- (291/(273+580)) = 65%. Blame thermodynamics.
So get 500MW for a week.

500MW for a week is enough to make a difference, however would the
capital be unused mostly? Yes, need a cross subsidy, but is scaleable.
The bigger the better to avoid thermal losses. The general idea is to
have a hot side of dry sand and plant equipment which moves in and out
of the dry sand. As far as I know sand goes a bit liquid if vibrated,
so shouldn't be a problem getting equipment in - with say a 20 x 30 x 10
meter pipe array with fins. However there is an easier solution which
is to lift and tip 8000 tons of hot sand at a time onto the heat exchanger.

The issue here is erosion of the pipework, when moving the sand, so
likely needs to be hardened, but the heat exchanger is accessible and so
easy to inspect and fix. Heating could by induction, hopefully iron is
cheap enough to have lots of of it, and leave heating elements in place.
Also electric heating could be high voltage AC.

It may be that a big deep hole is better to put the sand into, so really
constructing a geothermal cell. The problem is if it is too big how
does the pipework get fixed? Don't have joins and use a continual run.
Doubt you can get plastics which take high enough temperatures to get
the efficiencies you need.

Conclusion grid storage at scale in dry sand is possible and
economically feasible depending on the input electricity cost.

[Dave]

I had think about moving the sand into the heat exchanger, which would
be easy by letting it flow down, and again from the heat exchanger.
However getting it back up again isn't so easy. Balance weights might
help, so don't need massive motors or bucket lifts. An interesting
solution to dream with. Could use a rotating drum or similar, but again
it gets too complicated to construct so many of them with attachment
points.


Possibly best to find piping for laying in the sand and making sure it
is managed and compartmentalised. However as mentioned this affects the
scaleability as the pipework would only be used very occasionally e.g. a
few hours a week.

[mitchr...@gmail.com]

There is a claim of better batteries.
The industry ought to make them affordable.
It should provide that instead...

[edpr...@gmail.com]

Lots of ways to do heat storage.
https://en.wikipedia.org/wiki/Thermal_energy_storage
BTW, you probably don't want to pump sand. Use a fluid. Water may be a good choice.

Enjoy,
ed

[mitchr...@gmail.com]

How do you stop heat from coming out of its atom?

[Dave]

I didn't think you could get water hot enough for the efficiencies you
need for electricity generation, and I thought high pressure tanks are
expensive.

Moving sand has challenges. The Australians should know about this.
Regular Conveyor belts wouldn't be OK since the low temperature would
still be over 200 Celsius. Not sure about vortex pumps for sand. As
mentioned moving sand down should be easy, but putting it back up takes
energy. Any balance weights etc are a faff.

From the Wikipedia article, the salt tanks are at scale,
https://en.wikipedia.org/wiki/Thermal_energy_storage#/media/File:Abengoa_Solar_(7336087392).jpg
Wouldn't the marginal cost of salt be higher than sand? (10s of millions
of tons needed.)

I do have space at home for local heat storage (say 12m^3), and phase
change liquids are a lot more energy dense. It may make sense if I can
use free electricity to heat this. Not everyone in the UK is as lucky.

Also if hot enough may be able to get direct hydrogen production, but as
far as I know high temperature engineering has new challenges. It might
be better to produce hydrogen as needed and use at low pressure, and not
pressurise and store.

[Sergio]

energy transfer between two types of masses like that is at best 50%
efficient. [heat exchanger equations]

cheaper to pump water uphill storage into a lake

[Dave]

Did some new reading about this topic. Apparently Libyan sand is nice
round crystals, and mostly quartz. Sand with lots of limestone grains
wouldn't be a good, since it could decompose.

Correct as necessary. A million tonnes or a billion kilos of sand with
a heat capacity of 700J/ (kg K). Temperate drop from 800C to 120C total
650C drop gives in theory 650x700x1,000,000,000 J of heat.
= 455 x10^12 J.

Efficiency = 1 - Tc/Th (from wikipedia for realistic values)
= 1- 393/1073
= 0.63
or at most 50 % in a real world.

Thinking that engineering with steel etc might be more challenging with
higher temperatures.

Generated energy = 227.7 x 10^12 J
or 227700 GJ.
or 63.25 GWh.

This is enough to make a difference on a grid scale.

Best cost of a lithium ion battery can be 100USD /KWh
so to store 64GJ would cost

1 joule cost (Lion) 100/(1000x3600) = 27.77 x10^-6 USD
so 227700 GJ costs USD 6.323 Billion

This is quite a big investment, and not very scaleable because of the
cost.

A million tons of sand could cost USD 40 million.

Say four times the cost for building and engineering, so you have a
64GWH battery for about USD 200 million.

So the idea that hot storage, even with losses from thermodynamic
inefficiency, is worth further consideration holds up.



---

What size of building would be needed for million tons of sand?

Density about 1440 kg/m^3
so 1,000,000 tons is 694.444 m^3
Need a storage of about 200mx100mx36m
My understanding is that buildings under 40m are easier because they
shouldn't need aircraft lights etc.

[Dave]

There isn't enough geography like that in most countries. I worked with
someone who found a site like that and there are limited options
everywhere.

[Dave]

On 29/12/2020 12, Dave wrote:
>
...
The latest thinking has changed from moving the sand to a heat exchanger
where there is coolant flowing around, to dropping/injecting water into
the sand directly and making a lot of steam.

However the concept is meant to be that the bulk storage is really cheap
- a well insulated building or pit in the ground, which can be expanded
at build cost. If it needs to be able to take pressure the building
might need to be substantial. Why cheap expansion? This is meant for
storage for over three months, so the bigger the better, and keep costs
low.